isolated gate drives / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

isolated gate drives

previous 1 2 3 4 next last

Move Thread

LAN_403

HV Enthusiast

Thu May 11 2006, 03:05PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

blackplasma wrote ...

Hey, EVR, good to see someone looking at my lonely lil thread here.

Perhaps you could utter a few words of guidance regarding my [strange] ideas... ^^

Do you have a schematic to post???

Steve Conner

Thu May 11 2006, 03:57PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

OMG, EVR, are they diamond? Or kryptonite?

Marko

Thu May 11 2006, 04:55PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

As I could figure out reverse voltage is the main reason why schottkyes haven't pwned 'ordinary' diodes yet. For what I know 600V is the highest voltae silicon-carbide diodes acheived, and these still seeh relatively hard to get, and they don't have some great current rating (about 12A at most)

HV Enthusiast

Thu May 11 2006, 06:15PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

They are diamond based. And basically they are prototypes only provided to us for an R&D effort we are working on.

ragnar

Fri May 12 2006, 02:17AM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

EVR, here's a schematic...ish.

Wilson

Fri May 12 2006, 09:05AM

Registered Member #78 Joined: Thu Feb 09 2006, 11:27AM
Location: Sydney, Australia
Posts: 133

There seems to be a small error in regards to the schottky/freewheel section on your schem.
The schottky in series with the source of the FET is supposed to have its cathode connected to the anode of the freewheeling diode

Effectively, what you've just done is to isolate both the body diode AND the freewheeling one using the schottky, meaning that something will probably so pop or smoke when you get reverse current

Anyways, using an independant gate driver for each FET isn't quite a bad idea. You can drive the gates alot harder, and makes paralleling mosfets a more feasible idea.
Keep up the good work

Wilson

Steve Conner

Fri May 12 2006, 10:13AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Ya, Wilson is right. You should also have at least one pull-down resistor connected across one of those 1N4148s. I would tend to put one ~470 ohm resistor from each UCC input to ground. It's just to stop the transformer winding floating around.

Bodydiod

I found this on Obi-Wan Burnett's site

ragnar

Fri May 12 2006, 03:11PM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

Ah yes, schottky/body diode etc... OOPS on my part.

I'm getting the general impression from the datasheet that I should have my low-impedance/bypass caps on the output side of the UCC...

I'l considering sacrificing my principles and placing them "across" the output rail on pin6/7...

Should be relatively neat.

Part Scavenger

Sun May 14 2006, 02:04AM

Registered Member #79 Joined: Thu Feb 09 2006, 11:35AM
Location: Arkansas
Posts: 673

One problem. I don't see an audiomodulation circuit.

Looks pretty cool, and the layout is just sexy!

ragnar

Wed May 17 2006, 10:17AM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

I seemed to have had this board all perfect and nice, but had to mangle it by fitting that power schottky in there (to bypass the FET's body-diode)...

I've taken the zeners off the board because they were a PITA, and it's reasonably easy to clip them onto the MOSFET nice and close/directly. The gate resistor now has to go under the board because of the schottky being in the way. I'm not happy with the enclosed loop area between OUT1/SOURCE/GATE/OUT2, but it'll do.

I've got Steve-Conner-memorial schottkies on the outputs of both UCCs, and signal diodes on the inputs to clamp my +/-5V signal just in case it rings to something silly, and a 1K pulldown on each UCC input.

I've also put the power bypass caps on the output side of the chip and thickened up the MOSFET and DIODE and SCHOTTKY tracks.

I've got the UCCX732X datasheet here, and they're available in DIP8, SOIC8 and MSOP8... and funnily enough, the MSOP8 has a power dissipation rating almost 3 times that of DIP8 (1370mW versus
500mW)... it also has a C/W temp rating better than 10 times that of the DIP, when properly soldered down on its tinned thermal pad.

It's very tempting... USD$54 for 25ea of the MSOP UCC27322s (yes, TWO-7322 is rated for higher die temperatures than the 37322)...

They (the MSOP packages) would also be a much better fit when it comes to snuggling up to the FETs regarding loop area...

Has anyone got experience with the MSOP UCCs? Please rant here!

Any overhauls I should make to the layout before committing?

-Matt =)

Thanks mod, I forgot GIF's aren't resized.

previous 1 2 3 4 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob